PROJECT ABSTRACT/SUMMARY Fall prevention programs have had mixed results highlighting the need for more effective therapies that tackle the most common medical complication in stroke survivors. Task-specific training, which directly targets improving an individual's response during a fall, offers potential. The long-term goal of this proposal is to develop task-specific training that accounts for the unique balance challenges of stroke survivors that is both effective and time efficient. The ability to implement a task-specific training program in stroke survivors is thwarted by a lack of empirical evidence about which factors and mechanisms lead to a failed recovery during a balance disturbance in this population. An essential component of a successful recovery (avoidance of a fall) is compensatory limb movement ? taking a step to regain balance or reaching to a stable surface. We recently noted important similarities between compensatory movements and startle-evoked movements ? the involuntary release of planned movement when exposed to startling stimuli. Specifically, startle-evoked and compensatory movements share similar deficits following stroke ? delayed, diminished muscle activity, and inappropriate patterns of muscle activation. Thus our central hypothesis is that deficits in compensatory movements post-stroke are driven by deficits in startle-evoked movements. In Aim 1, our objective is to quantify stroke survivors' response during a balance disturbance identifying the key, trainable factors that lead to unsuccessful recovery (i.e. a fall). In brief, stroke survivors will be exposed to backward translations of a treadmill while quantifying stability measures, trunk and step kinematics, paretic limb parameters, and electromyography. Factors will be assessed for their ability to classify successful (fall avoidance) and failed (fall) recoveries. These factors can then be assessed for trainability in future task-specific training paradigms for stroke survivors. In Aim 2, we will pursue identification of the mechanism driving deficits in compensatory movements to arm perturbation in stroke survivors. Our objective is to evaluate the impact of deficits in startle- evoked movement on the ability of stroke survivors to effectively compensate during an arm perturbation. To precisely control perturbation parameters and allow bilateral assessment of paretic and non-paretic limbs, we will evaluate subjects while seated. In Aim 3, we will combine knowledge gained in Aims 1 & 2 and evaluate the link between factors that contribute to falls and the mechanisms proposed. Our objectives are to 1) evaluate the link between triggering a startle-evoked movement and successful compensatory movement during a balance challenge in unimpaired individuals and 2) evaluate the relationship between deficits in startle-evoked movement and deficits in compensatory limb movements post-stroke. The rationale for the proposal is to establish both the factors and mechanisms driving unsuccessful recovery during a fall post- stroke in order to generate an effective task-specific training paradigm in the next application.